CROSS REFERENCE TO RELATED APPLICATIONThis application is a divisional of U.S. patent application Ser. No. 12/614,883, filed Nov. 9, 2009 the entire content and disclosure of which is incorporated herein by reference.
BACKGROUNDThe present invention relates to the field of semiconductor structures, and particularly to a decoupling capacitor that employs a conductive through-substrate via and methods of manufacturing the same.
In resent years, “three dimensional silicon” (3DSi) structures have been proposed to enable joining of multiple silicon chips and/or wafers that are mounted on a package or a system board. The 3DSi structures increase the density of active circuits that are integrated in a given space.
As the circuit density increases unit area, the amount of switching activity per unit area also increases. This results in an increase in the noise generated on the reference supplies. As this noise increases, the performance of the internal devices as well as the performance of off-chip drivers is adversely impacted due to the reduction of noise margins available for the system design.
At present, this noise is controlled by embedding deep trench capacitors (DTC) within active silicon devices. To obtain sufficient degree of decoupling, a large array of DTC's are required. As the circuit density, switching activity, and power distribution structures are enhanced in a 3DSi structure, more DTC's will be required to control the noise generation. Further, as a number of DTC arrays are formed, there is an increase in the inductance between the active circuits and the arrays of DTC's, thereby requiring formation of additional DTC's to store the energy to be used to counter-balance a back electromagnetic force noise.
The voltage of the noise Vn is given by the following equation:
Vn=L×(dI/dt),
in which L is inductance, I is current, and t is time. As the amount of inductance (L) increases, or as the speed at which the current changes (dI/dt), which is proportional to the switching speed of circuits, the noise Vn increases proportionally.
The above considerations show that capacitive structures having low inductive is needed to control inductively noise generated within and transmitted into a 3DSi structure.
BRIEF SUMMARYAccording to an embodiment of the present invention, a capacitor in a semiconductor substrate employs a conductive through-substrate via (TSV) as an inner electrode and a columnar doped semiconductor region as an outer electrode. The capacitor provides a large decoupling capacitance in a small area, and does not impact circuit density or a Si3D structural design. Additional conductive TSV's can be provided in the semiconductor substrate to provide electrical connection for power supplies and signal transmission therethrough. The capacitor has a lower inductance than a conventional array of capacitors having comparable capacitance, thereby enabling reduction of high frequency noise in the power supply system of stacked semiconductor chips.
According to an aspect of the present invention, a semiconductor structure includes a semiconductor chip, which includes a semiconductor substrate; at least one capacitor embedded in the semiconductor substrate; and at least one laterally-insulated conductive through-substrate connection structure. Each of the at least one capacitor includes an inner electrode including a conductive through-substrate via (TSV) structure; a node dielectric laterally contacting and laterally enclosing the inner electrode; and an outer electrode laterally contacting and laterally enclosing a portion of the node dielectric.
According to another aspect of the present invention, a semiconductor structure includes a capacitor located in a semiconductor substrate and a contact structure located on the semiconductor substrate. The capacitor includes an inner electrode, a node dielectric, and an outer electrode. The inner electrode includes a conductive through-substrate via (TSV) structure that contiguously extends at least from an upper surface of the semiconductor substrate to a lower surface of the semiconductor substrate. The node dielectric laterally contacts and laterally encloses the inner electrode and contiguously extends from the upper surface to the lower surface. The outer electrode laterally contacts and laterally encloses a portion of the node dielectric. The contact structure is conductively connected to the outer electrode.
According to yet another aspect of the present invention, a method of forming a semiconductor structure is provided. The method includes forming a capacitor and a laterally-insulated conductive through-substrate connection structure in a semiconductor substrate. The laterally-insulated conductive through-substrate connection structure is formed by forming a dielectric tubular structure around a first through-substrate cavity formed in the semiconductor substrate; and filling a cavity within the dielectric tubular structure with a conductive material. The capacitor is formed by forming an outer electrode by doping a portion of the semiconductor substrate around a second through-substrate cavity; forming a node dielectric on a surface of the second through-substrate cavity; and forming an inner electrode by filling the second through-substrate cavity with the conductive material.
According to still another aspect of the present invention, a method of forming a semiconductor structure is provided. The method includes providing a semiconductor chip and electrically connecting the semiconductor chip to a mounting structure employing an array of solder balls. The semiconductor chip includes a semiconductor substrate; at least one capacitor embedded in the semiconductor substrate; and at least one laterally-insulated conductive through-substrate connection structure. The at least one capacitor has an inner electrode that includes a conductive through-substrate via (TSV) structure.
BRIEF DESCRIPTION OF THE SEVERAL VIEWS OF THE DRAWINGSFIGS. 1-18 are sequential vertical cross-sectional views through various processing steps of a first exemplary structure according to a first embodiment of the present invention.
FIG. 19 is a vertical cross-sectional view of a second exemplary structure according to a second embodiment of the present invention.
FIG. 20 is a vertical cross-sectional view of a third exemplary structure according to a third embodiment of the present invention.
FIG. 21 is a graph showing results of a simulation that shows a noise reduction at high frequency provided by an exemplary structure according to an embodiment of the present invention.
DETAILED DESCRIPTIONAs stated above, the present invention relates to semiconductor structures, and particularly to a decoupling capacitor that employs a conductive through-substrate via and methods of manufacturing the same, which are now described in detail with accompanying figures. Throughout the drawings, the same reference numerals or letters are used to designate like or equivalent elements. The drawings are not necessarily drawn to scale.
As used herein, a “conductive through-substrate via (TSV) structure” is a conductive structure that extends through a substrate, i.e., at least from a top surface of the substrate to a bottom surface of the substrate.
As used herein, a “laterally-insulated conductive through-substrate connection structure” is an assembly of a conductive TSV structure and another structure that laterally surrounds the conductive TSV structure and electrically isolates the conductive TSV structure from the substrate.
As used herein, a “mounting structure” is any structure to which a semiconductor chip can be mounded by making electrical connections thereto. A mounting structure can be a packaging substrate, an interposer structure, or another semiconductor chip.
As used herein, a first element “laterally contacts” a second element if there is a direct physical contact between the first element and the second element in a “lateral direction,” which is any direction perpendicular to a top surface or a bottom surface of a substrate.
As used herein, a first element “laterally encloses” a second element if an inner periphery of the first element is located on or outside an outer periphery of the second element.
As used herein, a first element “encapsulates” a second element if all outer surfaces of the second element are located within inner surfaces of the first element.
As used herein, two elements are “conductively connected” to each other if there exists a conductive path between the two elements to allow conduction of electricity.
Referring toFIG. 1, a first exemplary structure according to a first embodiment of the present invention includes asemiconductor substrate10 that has a semiconductor material. The semiconductor material of thesemiconductor substrate10 can be selected from, but is not limited to, silicon, germanium, silicon-germanium alloy, silicon carbon alloy, silicon-germanium-carbon alloy, gallium arsenide, indium arsenide, indium phosphide, III-V compound semiconductor materials, II-VI compound semiconductor materials, organic semiconductor materials, and other compound semiconductor materials. Preferably, the semiconductor material of thesemiconductor substrate10 is a single crystalline material. For example, thesemiconductor substrate10 can be a single crystalline silicon layer. Thesemiconductor substrate10 can be doped with dopants of a first conductivity type, which can be p-type or n-type. The dopant concentration of thesemiconductor substrate10 can be from 1.0×1014/cm3to 1.0×1017/cm3.
A dopedwell region12 is formed in thesemiconductor substrate12 by implanting dopants of a second conductivity through a portion of the top surface of thesemiconductor substrate12. The second conductivity type is the opposite of the first conductivity type. The second conductivity type is n-type if the first conductivity type is p-type, and vice versa. The dopant concentration of the dopedwell region12 can be from 1.0×1018/cm3to 1.0×1021/cm3to increase the conductivity of the dopedwell region12.
Referring toFIG. 2, apad dielectric layer16 and afirst mask layer18 are formed on the top surface of thesemiconductor substrate10. Thepad dielectric layer16 may, or may not, be formed on the backside of thesemiconductor substrate10. Thepad dielectric layer16 includes a dielectric material such as silicon nitride. Thefirst mask layer18 can be composed of a photoresist or a dielectric material such as silicon oxide or silicon nitride.
Referring forFIG. 3, thefirst mask layer18 is lithographically patterned, and the pattern in thefirst mask layer18 is transferred through thesemiconductor substrate10 by an anisotropic etch that employs thefirst mask layer18 as an etch mask. A first through-substrate cavity47 is formed in thesemiconductor substrate10. The lateral dimensions, e.g., diameter, a major axis, a minor axis, a length of a side, of the first through-substrate cavity47 can be from 1 micron to 100 microns, and typically from 3 microns to 30 microns, although lesser and greater lateral dimensions can also be employed.
Referring toFIG. 4, thefirst mask layer18 can be removed selective to thesemiconductor substrate10. A dielectrictubular structure20 is formed around the first through-substrate cavity47, for example, by converting exposed portions of thesemiconductor substrate10 on the sidewalls of the first through-substrate cavity47 into a dielectric material. For example, the exposed portion of the semiconductor substrate can be converted into a dielectric oxide by thermal oxidation. The dielectrictubular structure20 can include an oxide of the semiconductor material of thesemiconductor substrate10. For example, if thesemiconductor substrate10 includes silicon, the dielectrictubular structure20 can include silicon oxide. Thepad dielectric layer16 prevents conversion of other portions of thesemiconductor substrate10 into a dielectric material. The dielectrictubular structure20 extends from the top surface of thesemiconductor substrate10 to the bottom surface of thesemiconductor substrate10. A horizontal cross-sectional area of the dielectrictubular structure20 includes a hole corresponding to the first through-substrate cavity47. The thickness of the dielectrictubular structure20, as measured laterally between an inner periphery of the dielectrictubular structure20 and an outer periphery of the dielectrictubular structure20 can be from 100 nm to 1 micron, although lesser and greater thicknesses can also be employed.
Referring toFIG. 5, thepad dielectric layer16 can be removed. Optionally, adielectric liner30 is deposed on the inner sidewalls of the dielectrictubular structure20. Thedielectric liner30 can include, for example, a stack of a silicon oxide layer and a silicon nitride layer.
Referring toFIG. 6, the first through-substrate cavity47 is filled with a first disposable material to form a firstdisposable material layer49L. The firstdisposable material layer49L extends through thesemiconductor substrate10 and covers both sides of thesemiconductor substrate10, thereby encapsulating thesemiconductor substrate10. The first disposable material can be, for example, a polycrystalline silicon-containing material such as polysilicon or an amorphous silicon-containing material such as amorphous silicon.
Referring toFIG. 7, the firstdisposable material layer49L is removed from the front side and the backside of thesemiconductor substrate10, for example, by an etch-back process or chemical mechanical planarization (CMP). Further, a portion of the firstdisposable material layer49L is recessed below the top surface of thesemiconductor substrate10 by a recess depth rd, which can be from 200 nm to 2,000 nm, although lesser and greater recess depths rd can also be employed. The remaining portion of the firstdisposable material layer49L constitutes a firstdisposable material portion49.
Referring toFIG. 8, adielectric cap portion50 is formed by filling a cavity above the firstdisposable material portion49 with a dielectric material and removing excess dielectric material above a top surface of thedielectric liner30. Optionally, a silicon nitride cap layer (not shown) can be deposited on the top surface of thedielectric cap portion50 and the portion of thedielectric liner30 located on the front side of thesemiconductor substrate10.
Referring toFIG. 9, asecond mask layer51 is formed above the top surface of thesemiconductor substrate10. Thesecond mask layer51 can be composed of a photoresist or a dielectric material such as silicon oxide or silicon nitride. Thesecond mask layer51 is lithographically patterned to form an opening in an area that does not overlie thedisposable material portion49 or the dielectrictubular structure20. The opening in thesecond mask layer51 is formed over or in proximity to the dopedwell region12. The pattern in thesecond mask layer51 is transferred through thesemiconductor substrate10 by an anisotropic etch that employs thesecond mask layer51 as an etch mask. A second through-substrate cavity67 is formed in thesemiconductor substrate10. The lateral dimensions, e.g., diameter, a major axis, a minor axis, a length of a side, of the second through-substrate cavity67 can be from 1 micron to 100 microns, and typically from 3 microns to 30 microns, although lesser and greater lateral dimensions can also be employed.
Referring toFIG. 10, a dopedmaterial layer52 is deposited on the exposed surfaces of the first exemplary structure including the sidewalls of the second through-substrate cavity67. The dopedmaterial layer52 includes dopants of the second conductivity type. The dopedmaterial layer52 can be, for example, an arsenosilicate glass (ASG) layer. The thickness of the dopedmaterial layer52 is less than half of the smallest lateral dimension of the second through-substrate cavity67 to prevent plugging of the second through-substrate cavity67. Optionally, a dielectric capping layer (not shown) may be deposited over the dopedmaterial layer52 to prevent loss of dopants during a subsequent drive-in anneal.
Referring toFIG. 11, a drive-in anneal is performed to induce outdiffusion of dopants of the second conductivity type into a region of thesemiconductor substrate10 that surrounds the second through-substrate cavity67. An outer electrode is formed by doping a portion of thesemiconductor substrate10 around the second through-substrate cavity67. Specifically, theouter electrode60 is formed by converting a tubular region, i.e., a region in the shape of a tube, into a doped semiconductor region having a doping of the second conductivity type. For example, a dopant-containing material layer such as an arsenosilicate glass layer can be deposited on sidewalls of the second through-substrate cavity67 and the dopants can be driven into thesemiconductor substrate10 by a drive-in anneal. Theouter electrode60 is a doped tubular portion including a doped semiconductor material, i.e., has a shape of a tube. The lateral distance between the outer periphery of theouter electrode60 and the inner periphery of the outer electrode, i.e., the boundary with the dopedmaterial layer52, can be from 150 nm to 1,000 nm, although a lesser and greater lateral distances can also be employed. The dopant concentration of theouter electrode60 can be from 1.0×1018/cm3to 1.0×1020/cm3, although a lesser and greater dopant concentration can also be employed. The dopedmaterial layer52 is subsequently removed. In an alternate embodiment, theouter electrode60 can be formed by plasma doping without employing a dopedmaterial layer52.
Referring toFIG. 12, anode dielectric70 is formed on all exposed surfaces of the first exemplary structure including the inner sidewalls of theouter electrode60, which are the surfaces of the second through-substrate cavity67, and exposed surfaces of thedielectric liner30. Thenode dielectric70 is formed directly on sidewalls of the doped tubular portion while the disposable material is present in the semiconductor substrate. The thickness of thenode dielectric70 can be from 3 nm to 30 nm, although lesser and greater thicknesses can also be employed.
Referring toFIG. 13, the second through-substrate cavity67 is filled with a second disposable material to form a seconddisposable material layer77L. The seconddisposable material layer77L extends through thesemiconductor substrate10 and covers both sides of thesemiconductor substrate10, thereby encapsulating thesemiconductor substrate10. The second disposable material can be, for example, a polycrystalline silicon-containing material such as polysilicon or an amorphous silicon-containing material such as amorphous silicon.
Referring toFIG. 14, the seconddisposable material layer77L is removed from the front side and the backside of thesemiconductor substrate10, for example, by an etch-back process or chemical mechanical planarization (CMP). The remaining portion of the seconddisposable material layer77L constitutes a seconddisposable material portion77. The top surface of the seconddisposable material portion77 can be coplanar with a top surface of thenode dielectric70 on the front side of thesemiconductor substrate20.
Ahard mask layer72 is formed on one side of thesemiconductor substrate20, which is preferably the front side of the semiconductor substrate on which thedielectric cap portion50 is located. Thehard mask layer72 includes a dielectric material such as silicon oxide, silicon nitride, a doped silicate glass, or a combination thereof. The thickness of thehard mask layer72 can be from 500 nm to 5,000 nm, and typically from 1,000 nm to 3,000 nm, although lesser and greater thicknesses can also be employed.
Referring toFIG. 15, thehard mask layer72 is lithographically patterned to form openings over the seconddisposable material portion77 and the firstdisposable material portion49. Thedielectric cap portion50 is removed to expose an upper surface of the firstdisposable material portion49. An upper portion of the seconddisposable material portion77 can be removed during the removal of thedielectric cap portion50.
Referring toFIG. 16, the first dielectric material of the firstdisposable material portion49 and the second dielectric material of the seconddielectric material portion77 are removed by an etch that employs thehard mask layer72 as an etch mask. Removal of the firstdisposable material portion49 forms a cavity in a volume corresponding to the first through-substrate cavity47 in prior processing steps. This cavity is herein referred to as a re-formed first through-substrate cavity79, i.e., a first through-substrate cavity that is formed a second time Likewise, removal of the seconddisposable material portion77 forms a cavity in a volume corresponding to the second through-substrate cavity67 in prior processing steps. This cavity is herein referred to as a re-formed second through-substrate cavity78, i.e., a second through-substrate cavity that is formed a second time. The re-formed first through-substrate cavity79 is formed within the dielectrictubular structure20. Surfaces of thenode dielectric70 is exposed around the re-formed second through-substrate cavity78, and surfaces of thedielectric liner30 can be exposed around the re-formed first through-substrate cavity79. If thedielectric liner30 is not present, inner surfaces of the dielectrictubular structure20 can be exposed in the re-formed first through-substrate cavity79.
Referring toFIG. 17, the re-formed first through-substrate cavity79 and the re-formed second through-substrate cavity78 are filled with a conductive material to form a first conductive through-substrate via (TSV)structure80 and a secondconductive TSV structure82, respectively. The conductive material of the firstconductive TSV structure80 and the secondconductive TSV structure82 can include a doped semiconductor material, a metallic material, or a combination thereof. The conductive material of the firstconductive TSV structure80 and the secondconductive TSV structure82 can include, but is not limited to, doped polysilicon, a doped silicon-containing alloy, Cu, W, Ta, Ti, WN, TaN, TiN, or a combination thereof. The conductive material can be deposited, for example, by electroplating, electroless plating, physical vapor deposition (PVD), chemical vapor deposition (CVD), or a combination thereof.
After deposition of the conductive material, excess conductive material is removed from the top side and the bottom side of thesemiconductor substrate10 by planarization employing an etch-back process, chemical mechanical planarization, or a combination thereof. Top surfaces of the firstconductive TSV structure80 and the secondconductive TSV structure82 are coplanar with a top surface of thehard mask layer72. Bottom surfaces of theconductive TSV structure80 and the secondconductive TSV structure82 are coplanar with a bottom surface of remaining portions of the first exemplary structure. The bottom surface of the remaining portions of the first exemplary structure can be, for example, an exposed surface of thenode dielectric70 if a bottom portion of thenode dielectric70 remains after planarization or any other exposed surfaces at the bottom of the first exemplary structure. The firstconductive TSV structure80 and the secondconductive TSV structure82 are formed concurrently by employing the same deposition process and the same planarization process.
Referring toFIG. 18, acontact structure90 is formed by forming a trench through thehard mask layer72, thenode dielectric70, and thedielectric liner30 and by filling the trench with a conductive material such as a doped semiconductor material or a metallic material. Thecontact structure90 is conductively connected to theouter electrode60 through the dopedwell region12. The firstconductive TSV structure80, thenode dielectric70, and theouter electrode60 collective constitute acapacitor180, in which the firstconductive TSV structure80 is an inner electrode. The secondconductive TSV structure82, the portion of the dielectric liner contacting the secondconductive TSV structure82, and the dielectrictubular structure20 collectively constitute an laterally-insulated conductive through-substrate connection structure182. An end surface of the firstconductive TSV structure80, an end surface of the secondconductive TSV structure82, and an end surface of thecontact structure90 can be coplanar with an exposed surface of thehard mask layer72.
The first exemplary structure can be incorporated in a semiconductor chip. For example, a plurality of instances of thecapacitor180 and a plurality of instances of the laterally-insulated conductive through-substrate connection structure182 can be embedded in thesame semiconductor substrate10 of the semiconductor chip. The semiconductor chip may, or may not, include other semiconductor devices such as field effect transistors, bipolar transistors, thyristors, and diodes.
Eachcapacitor180 can include an inner electrode, which includes a first conductive through-substrate via (TSV)structure80, anode dielectric70, and anouter electrode60. The inner electrode contiguously extends at least from an upper surface of thesemiconductor substrate10 to a lower surface of thesemiconductor substrate10. Thenode dielectric70 laterally contacts and laterally encloses the inner electrode. Thenode dielectric70 contiguously extends from the upper surface to the lower surface. Theouter electrode60 laterally contacts and laterally encloses a portion of thenode dielectric70. Theouter electrode60 includes a doped semiconductor material.
The laterally-insulated conductive through-substrate connection structure182 includes a secondconductive TSV structure82 located in thesemiconductor substrate10 and a dielectrictubular structure20 laterally surrounding the secondconductive TSV structure82 and embedded in thesemiconductor substrate10. The laterally-insulated conductive through-substrate connection structure182 can include a portion of thedielectric liner30.
Referring toFIG. 19, a second exemplary structure according to a second embodiment of the present invention includes apackaging substrate200, a plurality offirst semiconductor chips100, a plurality ofsecond semiconductor chips300, an array offirst solder balls199 electrically connecting each of thefirst semiconductor chips100 to thepackaging substrate200, and an array ofsecond solder balls299 electrically connecting each of thesecond semiconductor chips300 to afirst semiconductor chip100. Each of thefirst semiconductor chips100 includes at least onecapacitor180 and at least one laterally-insulated conductive through-substrate connection structure182. Thefirst semiconductor chips100 may, or may not, include additional semiconductor devices such as field effect transistors, bipolar transistors, thyristors, and diodes. Thesecond semiconductor chips300 can include any type of semiconductor devices.
Thecapacitors180 can function as decoupling capacitors that reduce noise in a power supply system that supplies power to the devices in thesecond semiconductor chips300 and, if present, to the devices in thefirst semiconductor chips100. Eachcapacitor180 can provide a capacitance on the order of 1 pF to 10 nF, which is equivalent to the capacitance of 40-400,000 typical trench capacitors. Further, thecapacitor180 provides a lower inductance than a trench capacitor array that provides a comparable total capacitance. Thus, thecapacitors180 reduce noise in the power supply system especially during high frequency operations.
Referring toFIG. 20, a third exemplary structure according to a third embodiment of the present invention includes apackaging substrate200, ainterposer structure400, a plurality offirst semiconductor chips100, and a plurality of second semiconductor chips300. An array offirst solder balls199 electrically connects each of thefirst semiconductor chips100 to theinterposer structure400. An array ofsecond solder balls299 electrically connects each of thesecond semiconductor chips300 to afirst semiconductor chip100. An array ofthird solder balls399 connects theinterposer structure400 to thepackaging substrate200.
Theinterposer structure400 can include an interposerstructure substrate layer410, a lowerdielectric material layer420, and an upperdielectric material layer430. The interposerstructure substrate layer410 includes a plurality of through-substrate via structures that are schematically illustrated as vertical lines. The plurality of through-substrate via structures includes a plurality of capacitors180 (SeeFIG. 18) and laterally-insulated conductive through-substrate connection structure182 (SeeFIG. 18). The lowerdielectric material layer420 and the upperdielectric material layer430 can include metal lines that provide electrical wiring within the lowerdielectric material layer420 or the upperdielectric material layer430.
In general, a semiconductor chip including at least onecapacitor180 and at least one laterally-insulated conductive through-substrate connection structure182 can be mounted a mounting structure, which can be any structure on which the semiconductor chip can be mounted with electrical connections thereto. The mounting structure can be, but is not limited to, apackaging substrate200, aninterposer structure400, an assembly of aninterposer structure400 and apackaging substrate200, or another semiconductor chip such as asecond semiconductor chip300.
Referring toFIG. 21, a graph shows results of a simulation that shows a noise reduction at high frequency provided by an exemplary structure according to an embodiment of the present invention. The horizontal axis represents frequency of a noise component in a power supply system, and the vertical axis represents an equivalent impedance of a decoupling system including either a capacitor180 (SeeFIG. 18) according to an embodiment of the present invention or an array of trench capacitors according to prior art. The electrical noise in a power supply system is proportional to the equivalent impedance. The curve labeled “TSV w/582 pF” represents the equivalent impedance of acapacitor180 having a capacitance of 582 pF and constructed according to an embodiment of the present invention, e.g., as shown inFIG. 18. The curves labeled “DTC w/582 pF,” “2 nF,” and “4 nF” represent the equivalent impedance of trench capacitor arrays having a total capacitance of 582 pF, 2 nF, and 4 nF, respectively.
At a frequency range below 0.1 GHz, the voltage noise in the system power supply is limited by the total capacitance of a decoupling capacitor system. Above 1 GHz, however, the voltage noise in decoupling capacitor systems employing any of the trench capacitor arrays increases to with frequency on a converging curve irrespective of the total capacitance of the decoupling capacitor system because inductance of the decoupling capacitor system dominates. The decoupling capacitor system employing acapacitor180 of an embodiment of the present invention provides a lower voltage noise at frequencies above 1.2 GHz except for a small frequency range between 4 GHz and 4.5 GHz because thecapacitor180 has a low inductance. Thus, the decoupling capacitor system employing acapacitor180 of an embodiment of the present invention provides a superior performance in noise reduction while consuming less device area. In the second or third exemplary structure, if thefirst semiconductor chips100 do not include a semiconductor device, thecapacitors180 can be formed without requiring any area in the third semiconductor chips300. In the third exemplary structure, thecapacitors180 can be formed in a smaller area than an array of trench capacitors having a comparable total capacitance, thereby providing more area for other semiconductor devices that can be included in thefirst semiconductor chips100.
While the present invention has been particularly shown and described with respect to preferred embodiments thereof, it will be understood by those skilled in the art that the foregoing and other changes in forms and details can be made without departing from the spirit and scope of the present invention. It is therefore intended that the present invention not be limited to the exact forms and details described and illustrated, but fall within the scope of the appended claims.